Change the comment about sucky performance to recommendation to remove
DIAGNOSTIC/DEBUG from the kernel. This helped to boost the performance on my test machine from ~5KB/s to like 800KB/s for ftp test, and to like 700KB/s for netio test. I wouldn't have imagined the options are THAT costly ....
This commit is contained in:
parent
d4ede5e1e5
commit
2cd8833615
@ -1,4 +1,4 @@
|
|||||||
/* $NetBSD: if_ep_mca.c,v 1.3 2001/04/23 06:10:08 jdolecek Exp $ */
|
/* $NetBSD: if_ep_mca.c,v 1.4 2001/05/13 20:43:04 jdolecek Exp $ */
|
||||||
|
|
||||||
/*-
|
/*-
|
||||||
* Copyright (c) 2001 The NetBSD Foundation, Inc.
|
* Copyright (c) 2001 The NetBSD Foundation, Inc.
|
||||||
@ -70,15 +70,9 @@
|
|||||||
/*
|
/*
|
||||||
* Driver for 3Com 3c529 cards.
|
* Driver for 3Com 3c529 cards.
|
||||||
*
|
*
|
||||||
* Known issues:
|
* If you encouter sucky performance, try kernel without DEBUG/DIAGNOSTIC.
|
||||||
* - on my 386DX, speed of network is like 100KB/s at best; for bigger
|
* This helped on my test machine to change the performance of the card
|
||||||
* files fetched e.g. via ftp, I get as low as 5KB/s, mostly due
|
* from like 5KB/s to like 800 KB/s.
|
||||||
* to excessive number of overrun packets. Configuring system to
|
|
||||||
* use smaller TCP window might help, though the performance is
|
|
||||||
* still expected to be very dependant on CPU speed, especially
|
|
||||||
* since ISA and PCI attachments are reported to work OK on pentium
|
|
||||||
* class systems.
|
|
||||||
* Changing ic/elink3.c to use RX Early might help here potentially.
|
|
||||||
*/
|
*/
|
||||||
|
|
||||||
#include <sys/param.h>
|
#include <sys/param.h>
|
||||||
|
Loading…
Reference in New Issue
Block a user